Auditory hallucinations, experienced as external speech when none is present, are a cardinal symptom of schizophrenia. About 75 percent of patients with schizophrenia experience them at some point in their illness. To date most of what is known about hallucinations is based on verbal reports that are clouded by formal thought disorder or distrust of the examiners intentions. The study of the neurobiology of auditory hallucinations opens the possibility of developing objective indicators of this disabling symptom. Both hemodynamic and neurophysiologic approaches have identified the auditory cortex and language production and comprehension centers as being involved in auditory hallucinations. In addition, the middle temporal gyrus has been implicated in monitoring whether a voice is self- or other-generated. The investigator proposes to use both ERPs and fMRI to measure cortical activation. The investigator will compare schizophrenics who frequently hallucinate, those who rarely if ever do, and matched controls. Projects are designed to answer these broad questions: 1. Do brain areas involved in speech production and comprehension respond to speech and nonspeech sounds similarly in schizophrenics who hallucinate and those who do not? 2. Do areas of the brain involved in speech production, comprehension, and monitoring activate similarly in hallucinators and nonhallucinators during speaking aloud, listening to speech, and inner speech? 3. Are the same areas of the brain activated during speech and pitch discrimination in hallucinators and nonhallucinators?